We have a way to distinguish which query is a top level query. We are using an 8-node cluster of solr cloud, when any request is hit then min 9 and max 17 queries are being logged. It is consuming huge disk space. When we used to use solr standalone , log files used to be generated of ~30GB per day and now with solr cloud , the log file's size has increased to ~162GB.
On Thu, Sep 22, 2022 at 10:31 PM Shawn Heisey <apa...@elyograg.org.invalid> wrote: > On 9/22/22 09:17, Anjali Maurya wrote: > > Thanks Shawn for the suggestion. > > Can we make any change for the logging of only top level query not the > > shard level? > > I just tried doing a query on my tiny 9.1.0-SNAPSHOT SolrCloud install > that consists of one node, one shard, and one core, with ZK embedded. > The query was to the collection, not the core. I expected to see two > queries logged ... one for the collection and one for the core. I only > got one query logged, on the core. > > I know that when doing a sharded query that does NOT involve SolrCloud, > that both queries are logged. I am pretty sure that when both queries > are logged, it is not difficult to figure out which log entry is for a > top level query. > > Thanks, > Shawn > >
